Lakera Guard
Hosted detection of prompt injection, jailbreaks and data leakage for LLM apps.
Made by LakeraAI guardrails
About Lakera Guard
AI Agent Security protects agents across the full lifecycle, from discovery and risk assessment to runtime enforcement. It helps security teams understand which agents exist, what they can access, and whether their actions should be allowed or blocked in context.
The Lakera homepage states that it observes 100+ languages in real-time. The homepage quotes Dropbox saying it uses AI Agent Security to help safeguard its LLM-powered applications.
Drafted from lakera.ai · not yet confirmed by Lakera
How it works
Lakera Guard is an API you call on every prompt and response to detect prompt injection, jailbreak attempts, PII and secret leakage, and off topic or harmful content. It's fast enough to run inline and it's trained on a large set of real attacks, including the Gandalf game Lakera runs to collect them. Lakera was acquired by Check Point in 2025 and is being folded into its AI security line.
From the company
Reviews
Be the first to review it
AnonymousKnow a team weighing Lakera Guard? Send them this page.